@inproceedings{lei-etal-2023-unsupervised,
title = "Unsupervised Dense Retrieval with Relevance-Aware Contrastive Pre-Training",
author = "Lei, Yibin and
Ding, Liang and
Cao, Yu and
Zan, Changtong and
Yates, Andrew and
Tao, Dacheng",
editor = "Rogers, Anna and
Boyd-Graber, Jordan and
Okazaki, Naoaki",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2023",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.findings-acl.695",
doi = "10.18653/v1/2023.findings-acl.695",
pages = "10932--10940",
abstract = "Dense retrievers have achieved impressive performance, but their demand for abundant training data limits their application scenarios. Contrastive pre-training, which constructs pseudo-positive examples from unlabeled data, has shown great potential to solve this problem. However, the pseudo-positive examples crafted by data augmentations can be irrelevant. To this end, we propose relevance-aware contrastive learning. It takes the intermediate-trained model itself as an imperfect oracle to estimate the relevance of positive pairs and adaptively weighs the contrastive loss of different pairs according to the estimated relevance. Our method consistently improves the SOTA unsupervised Contriever model on the BEIR and open-domain QA retrieval benchmarks. Further exploration shows that our method can not only beat BM25 after further pre-training on the target corpus but also serves as a good few-shot learner. Our code is publicly available at \url{https://github.com/Yibin-Lei/ReContriever}.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lei-etal-2023-unsupervised">
<titleInfo>
<title>Unsupervised Dense Retrieval with Relevance-Aware Contrastive Pre-Training</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yibin</namePart>
<namePart type="family">Lei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Liang</namePart>
<namePart type="family">Ding</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yu</namePart>
<namePart type="family">Cao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Changtong</namePart>
<namePart type="family">Zan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andrew</namePart>
<namePart type="family">Yates</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dacheng</namePart>
<namePart type="family">Tao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2023</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anna</namePart>
<namePart type="family">Rogers</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jordan</namePart>
<namePart type="family">Boyd-Graber</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naoaki</namePart>
<namePart type="family">Okazaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Toronto, Canada</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Dense retrievers have achieved impressive performance, but their demand for abundant training data limits their application scenarios. Contrastive pre-training, which constructs pseudo-positive examples from unlabeled data, has shown great potential to solve this problem. However, the pseudo-positive examples crafted by data augmentations can be irrelevant. To this end, we propose relevance-aware contrastive learning. It takes the intermediate-trained model itself as an imperfect oracle to estimate the relevance of positive pairs and adaptively weighs the contrastive loss of different pairs according to the estimated relevance. Our method consistently improves the SOTA unsupervised Contriever model on the BEIR and open-domain QA retrieval benchmarks. Further exploration shows that our method can not only beat BM25 after further pre-training on the target corpus but also serves as a good few-shot learner. Our code is publicly available at https://github.com/Yibin-Lei/ReContriever.</abstract>
<identifier type="citekey">lei-etal-2023-unsupervised</identifier>
<identifier type="doi">10.18653/v1/2023.findings-acl.695</identifier>
<location>
<url>https://aclanthology.org/2023.findings-acl.695</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>10932</start>
<end>10940</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Unsupervised Dense Retrieval with Relevance-Aware Contrastive Pre-Training
%A Lei, Yibin
%A Ding, Liang
%A Cao, Yu
%A Zan, Changtong
%A Yates, Andrew
%A Tao, Dacheng
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Findings of the Association for Computational Linguistics: ACL 2023
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F lei-etal-2023-unsupervised
%X Dense retrievers have achieved impressive performance, but their demand for abundant training data limits their application scenarios. Contrastive pre-training, which constructs pseudo-positive examples from unlabeled data, has shown great potential to solve this problem. However, the pseudo-positive examples crafted by data augmentations can be irrelevant. To this end, we propose relevance-aware contrastive learning. It takes the intermediate-trained model itself as an imperfect oracle to estimate the relevance of positive pairs and adaptively weighs the contrastive loss of different pairs according to the estimated relevance. Our method consistently improves the SOTA unsupervised Contriever model on the BEIR and open-domain QA retrieval benchmarks. Further exploration shows that our method can not only beat BM25 after further pre-training on the target corpus but also serves as a good few-shot learner. Our code is publicly available at https://github.com/Yibin-Lei/ReContriever.
%R 10.18653/v1/2023.findings-acl.695
%U https://aclanthology.org/2023.findings-acl.695
%U https://doi.org/10.18653/v1/2023.findings-acl.695
%P 10932-10940
Markdown (Informal)
[Unsupervised Dense Retrieval with Relevance-Aware Contrastive Pre-Training](https://aclanthology.org/2023.findings-acl.695) (Lei et al., Findings 2023)
ACL